Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/arxiv-cornell-university/discus-the-deep-interior-scanning-cubesat-mission-to-a-rubble-pile-p13SBeCHdr
- ISSN
- 0273-1177
- eISSN
- ARCH-3330
- DOI
- 10.1016/j.asr.2018.06.016
- Publisher site
- See Article on Publisher Site
Abstract
We have performed an initial stage conceptual design study for the Deep Interior Scanning CubeSat (DIS- CUS), a tandem 6U CubeSat carrying a bistatic radar as the main payload. DISCUS will be operated either as an independent mission or accompanying a larger one. It is designed to determine the internal macro- porosity of a 260{600 m diameter Near Earth Asteroid (NEA) from a few kilometers distance. The main goal will be to achieve a global penetration with a low-frequency signal as well as to analyze the scattering strength for various di erent penetration depths and measurement positions. Moreover, the measurements will be inverted through a computed radar tomography (CRT) approach. The scienti c data provided by DISCUS would bring more knowledge of the internal con guration of rubble pile asteroids and their colli- sional evolution in the Solar System. It would also advance the design of future asteroid de ection concepts. We aim at a single-unit (1U) radar design equipped with a half-wavelength dipole antenna. The radar will utilize a stepped-frequency modulation technique the baseline of which was developed for ESA's technology projects GINGER and PIRA. The radar measurements will be used for CRT and shape reconstruction. The CubeSat will also be equipped with an optical camera system and laser altimeter to support navigation and shape reconstruction. We provide the details of the measurement methods to be applied along with the requirements derived of the known characteristics of rubble pile asteroids. Additionally, an initial design study of the platform and targets accessible within 20 lunar distances is presented. Keywords: Deep-space CubeSat, near earth asteroids, rubble pile asteroid, radar, computed radar tomography 1 Introduction initial stage mission concept, the Deep Interior Scan- ning CubeSat (DISCUS), in which the goal is to y two identical six-unit (6U) CubeSats (Figure 1) as The goal of this paper is to advance the mission de- a tandem into the orbit of a small rubble pile as- sign for recovering the deep interior structure of an teroid and to resolve its global interior structure via asteroid using small spacecraft. We introduce the arXiv:1805.01750v1 [astro-ph.EP] 4 May 2018 tomographic radar measurements. This is an impor- 2014; Rahmat-Samii et al., 2017; Hodges et al., 2016). tant scienti c objective which has been approached In Imken et al. (2017), the current status of 6U In- via several mission concepts (Safaeinili et al., 2002; terplanetary CubeSat development at the Jet Propul- Asphaug et al., 2003; Herique and Ciarletti, 2016; sion Laboratory (JPL) is given. Critical hardware, Kofman et al., 2007; Snodgrass et al., 2017). In par- such as data handling and communication systems ticular, we approach the Deep Interior concept by As- have been developed as standard buses at JPL. phaug et al. (2003) as a state-of-the-art small space- Namely, the Sphinx data handling system and the craft mission (NASA Mission Design Division, 2015) Iris transponder. The volume and weight of this which can be either independent or accompanying a hardware show that the promise of 6U interplan- larger one. DISCUS CubeSats will be equipped with etary CubeSats with around 2U housekeeping, 2U a bistatic penetrating radar and an optical imaging propulsion and almost 2U payload could be kept. system. The design is derived from airborne Ground An ESA/ESTEC study found the limitations of 6U Penetrating Radar (GPR) which is today applied in for a NEA mission to be critical, especially regard- mapping subsurface glacier and soil structures, e.g., ing on thermal management and v requirements ice thickness (Rutishauser et al., 2016; Gundelach of independent interplanetary missions. Therefore, a et al., 2010; Eisenburger et al., 2008; Fu et al., 2014). 12U based CubeSat concept M-ARGO (Walker et al., The rst realized tomography attempt for small a 2017) has been investigated and found promising. Be- Solar System body was the COmet Nucleus Sound- sides completely autonomous missions also so-called ing Experiment by Radio-wave Transmission (CON- CubeSat piggy-bag missions can become more com- SERT) (Kofman et al., 2007, 2015) during the Euro- mon in the future. In the piggy-bag approach, a pean Space Agency's (ESA) Rosetta mission in 2014. mothership transports the CubeSats to its target, de- In CONSERT, a radio-frequency signal was transmit- ploys them and serves as a communication link. The ted between the Rosetta orbiter and its lander Phi- advantage is that the requirements for navigation, lae. Rosetta was a major scale mission with its total communication, propulsion and total ionizing dose budget of 1.4 Billion Euro with the cost of the lan- (TID) are reduced. For example, the plans of the der being 200 Million. Even thought the missions unrealized Asteroid Impact Mission (AIM) included has been a success the high costs hinders the de- two CubeSats and the Mascot-2 lander (Michel et al., velopment of a comparable successor on this scale. 2016; Herique and Ciarletti, 2016). Due to lower cost and the recent technological ad- DISCUS (Figure 1) is designed to detect the in- vances, small spacecraft would enable the exploration ternal macroporosity of an Itokawa-size (Abe et al., of both asteroids and comets more exibly and in 2006) 260 to 600 m diameter Near Earth Asteroid higher numbers. For comparison, the recently pub- (NEA) from a few kilometers distance. The primary lished M-ARGO concept aims for a budget of 25 Mil- goal of DISCUS will be to achieve a global signal pen- lion Euro (Walker et al., 2017). The interest towards etration as well as to analyze the scattering strength using CubeSats to complement traditional planetary for various di erent penetration depths and measure- missions is constantly increasing and, therefore, mis- ment positions. Additionally, forming an actual 3D sion design to support this development is needed. reconstruction based on the measurements will be at- A few years ago, Staehle et al. (2013) published an tempted. Using the re ected wave, also the shape of initial report on potential Interplanetary CubeSats the body can be determined. Due to the many limi- with 2U science payload and 2U propulsion module. tations of space missions, e.g., the strict payload and Later on in 2014, NASA announced with NeaScout energy bounds, our radar design aims at a minimal the rst detailed mission design of a 6U CubeSat to weight and power consumption. Akin to airborne y with a solar sail to an asteroid (Frick et al., 2014). GPR, we apply the stepped-frequency measurement Recently developed CubeSats capable of operating in technique and a half-wavelength dipole antenna (Fu deep space include the Lunar IceCube (Clark et al., et al., 2014). Maximizing the detectability of deep 2016) and Mars Cube One (Asmar and Matousek, structures such as voids necessitates using a signal 2 Figure 1: Conceptual pictures of the 6U DISCUS design containing (a) 1U radar (green), (b) 1U Commu- nication (yellow), (c) 2U BIT-3 propulsion system (gray). Attached to the frame is a linear half-wavelength dipole antenna (yellow line). During the launch, the antenna will be rolled up. The other parts are com- ponents of the shelf (COTS). Among others a Pumpkin 6U frame with four star trackers, an altimeter, a camera, three reaction wheels and PCB boards for housekeeping and a cold gas thruster. Two solar panels are oriented in a perpendicular direction with respect to the antenna. The radar measurement is performed with the antenna pointed towards the sun in order to minimize the noise caused by solar radiation. It also maximizes the power gain of the solar panels and reduces the need for power storage. frequency bellow 100 MHz (Francke and Utsi, 2009; stage. Exploring their population is a key to under- Leucci, 2008; Kofman, 2012). We initially target at standing the Solar System's history and evolution, 20{50 MHz center frequency in the radar design. although they contain only about 0:002 % of its to- tal mass. The asteroids have evolved through several This paper is structured as follows. In Section 2, we processes; space weathering changed their surface by motivate the mission plan via a brief review of what e ects of radiation, internal heating may have altered is known of rubble pile asteroids. Section 3 describes their chemistry and mineralogy, and collisional events the on-board scienti c instruments and their use. We transformed their interior structure and size distri- provide the details of the measurement methods to bution. According to the present knowledge, most of be applied along with the requirements derived from the asteroids are actually resulting from catastrophic the known characteristics of the rubble pile asteroids. disruptions of large parent bodies (Farinella et al., Section 4 sketches the mission layout. An initial de- 1982). Namely, modeling of catastrophic disruptions sign concept of the platform and accessible targets is (Michel et al., 2001, 2002) can reproduce the size dis- presented. Section 5 is devoted to the discussion. Fi- tribution of the Karin asteroid family, showing that nally, Section 6 concludes the paper and summarizes many fragments re-accumulated to form rubble pile the study. asteroids. The current evidence also shows that these pro- 2 Rubble pile asteroids in the cesses have left most of the asteroids in the size range of about 200 m to 100 km with an interior that is not Solar System monolithic but rather an agglomerate of smaller frag- ments, bound together mainly by gravity and only When the Solar System was formed 4:6 Ga ago, weak cohesion (Richardson et al., 2002). planetesimals constituted the building blocks of pro- toplanets and eventually the planets themselves. The rough comparisons between the meteoritic ma- Comets and asteroids are the remainders of this early terials and the indirect density estimates suggest 3 that many asteroids have a large amount of inter- ments, explaining the enhancement of the large boul- nal macroporosity, that is, voids larger than the typ- ders on its surface. Namely, the amount of regolith ically micrometer sized cracks in the matrix of mete- observed on Itokawa's surface is signi cantly larger oritic samples (denoted as microporosity). Recently, than what could be produced by impact events form- Carry (2012) found high percentages of macroporos- ing the observed number of surface craters (Barnouin- ity by comparing mass and volume of 287 asteroids, Jha et al., 2008). Itokawa's rough highlands terrain is and aggregating information about meteoritic sam- covered with rubbles following a cumulative power ples linked to their taxonomic classes. For exam- law distribution of 3:1 0:1, and the number of ple, the estimate P = 72 14 % was obtained large boulders is enhanced compared to models of re- macro for asteroid (854) Frostia. The measurements pro- aggregation after a catastrophic impact event, which vided by the recent Hayabusa and Rosetta missions most likely formed the asteroid (Michikami et al., have also shown that the small asteroid Itokawa (Fig- 2008). Indirect inferences of the macroporosity struc- ure 2a, Saito et al., 2006; Abe et al., 2006) as well as tures can also be obtained through the simulation of the larger one Steins (Figure 2b) might have around the impact events which allows one to connect the 40 % of macroporosity, respectively. Moreover, simu- crater structure to the surface features and also to lations of granular owing material match well with determine the shock wave propagation characteristics the observed top-like shape of many observed aster- inside the asteroid. Deller (2017) demonstrated this oids, e.g., Steins; simulating the spin-up of rotating method on asteroid Steins, showing that the catena asteroids, Walsh et al. (2008) as well as S anchez and of pits running from the rim of the crater on Steins Scheeres (2012) found a material ow towards the can be interpreted as a sequence of sinkholes into an equator, forming asteroids with comparable shapes. interior crack, which has opened due to the impact. Analyzing the rotation period as a function of the The mission proposed here provides the means to size has revealed that, for asteroids larger than 200{ nally sample the interior of a rubble pile asteroid 300 m, there exists a general cut-o at a period of directly and to provide direct measurements about 2:2 hours which only some so-called superfast rota- the distribution of porosity inside these objects. The tors such as the asteroid 1950 DA surpass (Pravec knowledge obtained would be vital, for example, for et al., 2002; Rozitis et al., 2014). This spin barrier the assessment of the asteroid impact de ection as a (e.g. Pravec and Harris, 2000; Kwiatkowski, 2010) is possible way to neutralize threats by potentially haz- coinciding with the limit for a cohesionless body be- ardous earth colliding asteroids. Namely, the inter- fore it starts loosing mass at the equator, which in- nal structure is of crucial importance in this regard, dicates that most bodies larger than 200{300 m in since the scale of the porosity inside asteroids deter- diameter are not monolithic in the interior. mines the propagation of shock waves and, therefore, the eciency of energy transfer during impact events. While the evidence for a shattered interior of aster- Another important application would, obviously, be oids is strong, the exact con guration of the interior the mineralogy of asteroids, since the proposed radar of these objects is unclear, as direct measurements measurement technique would provide rst-hand in- are still missing. Macroporosity can arise from void formation of the dielectric properties in the interior space inside the asteroid at various di erent scales. (Michel et al., 2015). There might be large void fractures at the scale of NEAs that are easy to reach for space missions hundreds of meters in the interior, as shown to be have been explored, e.g., in the Mission Accessi- likely for Steins (Figure 2b, see details in Deller, ble Near-Earth Objects Survey (MANOS, Thirouin 2017), or small scale cavities caused by interlock- et al., 2016) focusing on NEAs that have orbital pa- ing of boulders. Moreover, the sorting e ects of the rameters with low v, i.e., low relative velocity with granular mechanics suggest that the asteroid Itokawa respect to the Earth. For a small, but well-described might be a rubble pile with large constituent frag- sample of 86 sub-kilometer sized asteroids, it rated https://cneos.jpl.nasa.gov/doc/1950da/ seven of these objects as possible targets for manned (a) Image of the asteroid Itokawa taken by the Hayabusa (b) Asteroid Steins imaged by the OSIRIS camera sys- spacecraft. Source: JAXA, http://global.jaxa.jp/ c tem on board of Rosetta. Source: ESA 2008 MPS article/special/hayabusa_sp3/index_e.html for OSIRIS Team MPS/UPD/LAM/IAA/RSSD/INTA/ UPM/DASP/IDA Figure 2: Itokawa and Steins, two rubble pile asteroids visited by spacecraft. missions. Using expectations on the number of NEAs from de-biasing studies of the NEA population (Tri- carico, 2017), Thirouin et al. (2016) estimate that a total number of approximately 33 000 NEAs are accessible for space missions in the Thirouin et al. (2016) de nition. 3 Scienti c instrumentation 3.1 Stepped-frequency radar The radar design of DISCUS is based on the stepped- frequency technique which requires a relatively low DC power and low data rate. This is possible since, Figure 3: Stepped-frequency radar for the HOPE instead of a single full-bandwidth signal pulse, narrow (Handheld Operated Demining System) (RST Radar frequency bands are transmitted and received sep- Systemtechnik AG, 2001). The dimensions of the box arately. As the stepped-frequency technique allows are 12 cm 15 cm 5 cm and its volume is 900 cm . measurement of weaker bands, it is generally used in radars with a narrow instantaneous bandwidth as compared to the total bandwidth of the resolution aimed (Lacomme, 2001). 5 3.1.1 Mathematical concept cies (Kraus, 1966). The minimum total e ect of the active Sun and the galactic background corresponds The set of transmitted narrow frequency lines to about 205 MHz, since the galactic noise increases ; ; : : : ; allows one to approximate a given 1 2 N towards the low frequencies. In order to minimize the function f via the sum f = c . If ' is the ` ` ` `=1 e ect of the Sun, the antenna will be pointed towards received signal (data) corresponding to frequency line it during the measurement. The targeted radar spec- , then the data g resulting from a transmission f i cations have been listed in Table 1. is approximately given by g = c ' , where the ` ` `=1 The duration of a single frequency line can be es- coecients c ; c ; : : : ; c follow from the formula for 1 2 N timated via = 1=B , where B = B=N , i.e., the to- ` ` f . Hence, the data g for any transmission f within tal bandwidth B is divided by the number N of the the given frequency range can be approximated, if the frequency lines, roughly around 64 lines with pulse function pairs and ' for ` = 1; 2; : : : ; N are given, ` ` duration 32 μs can be transmitted and received sep- i.e., if the stepped-frequency measurement data are arately from a 5 km distance to the asteroid (the available. For a mathematical study of the DISCUS pulse duration is 96 % of the total travel-time). A radar, please refer to Takala et al. (2017a). sucient pulse repetition time (PRT) with respect to the corresponding decay time will be 150 μs, i.e., 3.1.2 Baseline design the duration of the complete pulse sequence will be about 9:6 ms at the 5 km distance. The narrower is The stepped-frequency radar design referred to in B the longer is the single pulse duration and the this study was originally developed and veri ed by more energy is transmitted. Even when assuming at RST Radar Systemtechnik AG within the European worst a ground speed of a few meters per second, a Space Agency's (ESA's) GINGER (Guidance and sucient number frequency lines could be measured. Into-the-Ground Exploration Radar) project (Lentz Lower speed would hereby enable to transmit and re- and Braun, 2000; Putz et al., 1997), where a plane- ceive more lines and enhance the signal-to-noise ratio. tary rover was used as the reference platform. In a The stepped-frequency sequences for larger measure- subsequent ESA project PIRA (Planetary Into-the- ment ranges can also be longer in time and, thereby, Ground Radar and Altimeter) this technology was contain more lines in order to extend the ambiguity veri ed under platform movement using a helicopter and dynamic range. For each frequency band, the as the carrier (Braun et al., 1997). In addition to spectrum of the signal is recorded for a time interval space systems, RST has also previous experience on determined by the observation distance. packing a stepped-frequency radar into a very small volume. For example, in the HOPE (Handheld Oper- ated Demining System) landmine detector the total 3.1.4 Link budget and measurement noise radar volume was 900 cm (Figure 3). The initial link budget (Table 2) for the received sig- nal is here approximated in decibels via the equation 3.1.3 DISCUS radar P = P + G + S + G where P = 0 RX TX TX RX TX dB (10 W) and P denote the transmitted and re- In the 6U DISCUS design, the goal is to t a radar RX ceived power, S the signal power at the receiver lo- with a 2 MHz total signal bandwidth and 20{50 MHz cation based on an isotropic radiator model, with center frequency into a 1U cube (1000 cm ). The an isotropic source and e ective antenna aperture lower end of the targeted frequency interval is ad- vantageous in order to achieve an appropriate signal A = =(4), and G and G the gain of the eff TX RX penetration (Kofman, 2012) and to minimize solar transmitting and receiving half-wavelength dipole an- noise. Namely, if the Sun is active, the dominat- tenna, that is, (G = G = 2:15 dBi). TX RX ing error source will be the solar radiation (see Fig- In order to obtain an estimate for S , we sim- ure 4) which is on a comparably low level at the tar- ulated the planned radar measurement numerically geted interval decreasing towards the lower frequen- (Takala et al., 2017a) using a 550 m diameter syn- 6 Table 1: Targeted radar parameters. Center frequency: 20{50 MHz Antenna length: Two times 1.5{3.75 m (=2) Radar modulation: 32 to 2048 lines Pulse duration (5 km distance): 32 μs Pulse repetition time (5 km distance): 150 μs Input/ Transmitting power: 40 W/ 10 W Receiver bandwidth: 2 MHz Size: 1 U Weight: 1 kg thetic asteroid model based on the surface shape of distance amount to a total of roughly around 7{15 1998 KY26 (Figure 4). The center frequency of the dB using the galactic background noise as a refer- measurement was assumed to be 20 MHz, the mea- ence. Due to the two-way signal path and the linear surement distance 5 km and the number of frequency distance dependence of the (stepped-frequency) pulse lines N = 64. The relative dielectric permittivity length, approaching the asteroid from the distance d distribution " enclosed by the surface consisted of to d would increase P in decibels approximately r 1 RX a background part (" = 4), a 40 m thick surface according to the formula 30 log (d =d ) r P 0 1 RX 10 layer (" = 4) and three 60 to 100 m diameter voids which gives +5 dB for d = 3:5 km and r P 1 RX (" = 1) located in the deep interior part. The av- +9 dB for d = 2:5 km. r P 1 RX erage signal attenuation rate was around 25 dB=km corresponding to a porous basalt at the chosen fre- 3.1.5 Computed radar tomography quency (Kofman, 2012). The peak power P re- RX ceived for the surface layer and a void in the depth Tomographic analysis of the targeted NEA can be of 150 m was 120 dB and 128 dB, respectively. performed akin to ordinary tomographic GPR sur- veys in which radio frequency signals enable re- The relative measurement noise levels (Table 2) constructing subsurface ground layers. Based on caused by the spectral radiation ux density F the radar speci cations, the range resolution ( = (Kraus, 1966) from the active and quiet Sun (2E- r c=(2B) with c denoting the signal velocity and B the 19 and 2E-23 W/m /Hz) as well as from the galac- bandwidth) determining the minimum distance be- tic background (5E-20 W/m /Hz) were estimated tween two separable details is 20 to 40 m inside the via the formula 10 log (FB A =P ) in which ` eff TX NEA, assuming that its relative electric permittivity A = 0:13 is the e ective aperture of the half- eff is 3 to 12 (e.g. solid and pulverized rocks and silica). wavelength dipole antenna, B the bandwidth of a On ground, the numerical simulations necessary for single frequency line, and P is expressed in Watts, TX inverting the radar measurements can be performed respectively. Our experience is that the data can be using a state-of-the-art computer or computing clus- inverted when the total measurement noise level is at ter with sucient memory for running iterative wave least 5{10 dB below the peak level of P (Takala RX equation solvers within a three-dimensional compu- et al., 2017a,c). In (Jol, 2008; Erst, 1984), -8 dB tation domain. Our recent simulation study demon- has been suggested as the maximal tolerable noise strates that a 2 MHz full-wave (full-bandwidth) data level for penetrating radar measurement. That is, can be computed and inverted for the targeted aster- losses related to penetration, hardware, and signal oid size range using a high-end workstation computer processing, and miscellaneous reasons can at 5 km equipped with up-to-date graphics processing units https://ssd.jpl.nasa.gov/sbdb.cgi?sstr=1998KY26 (GPUs) (Takala et al., 2017a). 7 Table 2: Initial data link budget and noise level estimates for 5 km orbiting distance, 20 MHz center frequency, 2 MHz bandwidth, and 10 W transmission power. Transmitted power P 0 dB TX Half-wavelength dipole antenna gain G = G 2:15 dBi TX RX Peak power P received for the surface layer 120 dB RX Peak power P received for a void at 150 m depth 128 dB RX Measurement noise due to active Sun 137 dB Measurement noise due to quiet Sun 177 dB Measurement noise due to the galactic background 143 dB Figure 4: Left: The surface model of the asteroid 1998 KY26 scaled to a diameter of 550 m to examine the relative received power P for a 10 W signal pulse with 20 MHz center frequency and 2 MHz bandwidth RX (Takala et al., 2017a). Right: The backscattered signal power P (dB) received by the transmitting RX spacecraft at 5 km orbiting distance to the asteroid. The peaks re ecting from a surface layer and a void at 150 m depth have been marked with vertical lines. The noise levels corresponding to active and quiet Sun and the galactic background have been illustrated with horizontal lines. Also, the maximal theoretical noise tolerance -8 dB suggested for a penetrating radar signal (Jol, 2008; Erst, 1984) is shown. The x-axis shows the propagation time of the signal (s). 8 To record the tomography data, we will use a sim- 3.3 Optical imaging ple bistatic (two-spacecraft) measurement scenario in A visual camera system is needed to enhance the ac- which the CubeSats will be placed within a (nearly) curacy of spacecraft positioning. The images pro- constant angle from each other as seen from the vided by the camera can also be used to optimize NEA. One of the CubeSats will both transmit and the shape model. Because of the limited data trans- receive the signal and the other one will serve as an mission rate and the resulting constrains on the data additional receiver. As the instrumentation on both volume, the angular resolution of the camera has to CubeSats is identical the concept is redundant by its be chosen to provide a surface coverage with a reason- design. The reference phase can be transmitted by able amount of images in the expected orbit. With- the master satellite. The ying formation does not out detailed analyses available at this point, we build necessarily need to be maintained stationary: the mu- our estimate on the experience from the Hayabusa tual distance between the CubeSats can slightly vary mission (Ishiguro et al., 2010). The AMICA imag- during the measurements, since their positions will ing system (Nakamura et al., 2001) on the Hayabusa be tracked independently. At least the following two spacecraft took a total of > 1400 images of asteroid di erent choices for will be used: (i) 35 and Itokawa. The camera had a eld-of-view (FOV) of (ii) 145 . The rst one of these (i) will enable an 5:83 5:69 , and a detector with 1024 px 1000 px. enhanced backscattering measurement; our numeri- The analog-to-digital converter sampled the signal cal results (Takala et al., 2017a) suggest that with with 12 bit resolution. This corresponds to an im- the choice = 25 the accuracy and reliability of the age le size of 1:46 MiB uncompressed. Lossless com- bistatic tomography outcome is superior compared pression can achieve an estimated factor 2 in le size to the monostatic (single-spacecraft) case. The sec- reduction. Therefore, we expect to need to downlink ond option (ii) will allow a higher signal-to-noise ratio a total of 1025 MiB over the course of the mission due to one-way signal paths and tomographic travel- in order to achieve a shape model for a scenario sim- time measurements, performed, e.g., in CONSERT ilar to Hayabusa. Commercial of the shelf camera (Takala et al., 2017b; Kofman et al., 2015). systems mounted on small satellite platforms achieve a FOV of about 15 . While the lower resolution di- rectly decreases the resolution achievable for the sur- face shape model, the estimate of the data volume is 3.2 Shape model still applicable. The altimeter-based shape model of the asteroid can be reconstructed via a combined use of radar and 4 Mission layout laser measurements. The laser data will complement the radar measurements in which the rst return 4.1 Overall design coming from the surface of the asteroid is detected. Generally, the topographic ( at surface) height accu- For the 6U design of DISCUS, mostly components of racy is below 20 % of the range resolution (Man- th the shelf (COTS) have been selected. The spacecraft asse, 1960). Consequently, in empty space, a mea- is illustrated in Figure 1. The frame is based on the surement with the bandwidth of 2 MHz has the range 6U Supernova Bus of Pumpkin Inc. Electric propul- resolution of = 75 m and height accuracy of about sion will be used. An COTS engine that can provided = 15 m. These values are independent of the op- th the required v within the power and mass require- eration distance. The latter one serves as a rough 3 ments is the Busek BIT-3 . The engine is also used estimate for the shape model reconstruction accu- for other CubeSat missions such as the Lunar Ice- racy assuming that a suciently dense distribution Cube (Folta et al., 2016). With a wet weight of 3 kg of points over the asteroid surface can be covered at the orbit. http://www.busek.com/technologies__ion.htm 9 5 including 1:5 kg of propellant, the BIT-3 delivers a BOL peak power . The system consists, hereby, of v of 3:2 km s for a 13 kg CubeSat, when operated two triple folded 2U 3U solar panels each mounted at 75 W power. on one gimbaled root hinge. The power consumption estimates have been budgeted in Table 3. A Vacco cold-gas thruster and four star trackers complement the COTS attitude determination and control system. For power generation, the solar panel 4.3 Antenna system MMA eHAWK with 96 W beginning-of-Life Commercial systems using a motor have already (BOL) has been chosen. The camera and a mili- proven the feasibility of a 3 m deployable antenna for tary rated JenOptic altimeter are mounted in the Cubesats. In example Oxford Space ASTROTUBE front next to the dipole antenna. Additionally 1U BOOM has been validated in space, by unfolding to is occupied by housekeeping systems such as battery 1:5 m (Reveles et al., 2017). The Oxford system is systems and an on-board computer. The green and available for length of up to 3 m and ts into 1U at yellow boxes represent the radar and communication a weight of 0:6 kg. Our half-length dipole antenna system, respectively. The estimated weight of these is projected to be build based on a metal strip de- systems leads to the total wet mass of the satellite, sign, which is widely utilized for antennas in space with 20 % margin of around 13:2 kg. The mass bud- applications. A self-enrolling strip has been chosen get can be found in Table 3. as the design basis instead of a motor in order to Since the bistatic measurement process requires ac- reduce mass and volume of the system. The struc- curate timing, e.g., regarding radar pulse window ture is slightly bent to be more stable and to function synchronization, a chip-scale atomic clock (CSAC) as a pre-stressed spring when rolled up. A thermal will be needed on board in order to minimize all pos- knife release the tapes, which unwinds entirely by sible clocking di erences between the two CubeSats rolling of the spacecraft. First unfolding tests have (DeNatale et al., 2008). The synchronization accu- been successful. Currently, the stability during orbit racy provided natively by the CSAC will be around maneuvers is under investigation. 100 ns , that is, the travel-time corresponding to the maximal attainable radar accuracy (20 % of the range resolution, Section 3.2). The precision of the CSAC 4.4 Targets will allow a several hours' measurement time without In addition to the scienti c and instrument require- re-synchronization, i.e., control pulse exchange. ments, the target needs to be reachable for a Cube- Sat. Based on a lunar transfer orbit, a study for 4.2 Power requirement an asteroid encountering mission with a 6U CubeSat has been performed by Landis et al. (2014). Landis In the science phase, the total peak power consump- 1 argues that at least an initial v of 1:6 km s into tion will be about 50 W. The stepped frequency radar various asteroid orbits could be provided from the will require 40 W non-continuously in short pulses. lunar transit insertion. Thus, the maximum possi- The attitude control and housekeeping need together 1 ble total v will be 4:8 km s , i.e., the sum of the up to 10 W. During the cruise phase, the electric 1 v given by the thruster (3:2 km s ) and that of propulsion system requires most of the power. The the lunar transit. As the detailed orbit analysis is desired Busek BIT-3 engine can be operated between 1 still missing, we assume a margin of 0:8 km s and 55 and 80 W. The eciency of the engine will de- limit our survey to asteroids with a v of maximally crease along with the power consumption. 1 4 km s . Still, various asteroids ful ll this require- The MMA eHAWK is supplying solar arrays in the 6 ment. Figure 5 shows targets on their closest ap- required scale. The HaWK 112-7058 delivers 96 W proach that are reachable on various launch dates. 4 5 https://www.microsemi.com/document-portal/doc_ http://mmadesignllc.com/existing-hawk-configurations/ view/133866-low-noise-csac-datasheet https://cneos.jpl.nasa.gov/ca/ 10 Component Speci cation Availability Mass (kg) Power (W) Bus 6U w. shielding modif. COTS 2:0 Solar panels w. pointing announced 1:0 96:0 Propulsion Bit 3 COTS 3:0 56:0 to 80 Thruster Cold gas, 50 Ns COTS 0:6 0:0 to 10:0 Attitude control 2 mNm COTS 1:0 0:5 to 2:0 Housekeeping Incl. thermal modif. COTS 1:0 2:0 to 7:0 Radar TRL 3 1:0 40:0 Dipole antenna 2 3:75 m TRL 2 0:4 Atomic clock Chip-Sized Atomic Clock COTS 3:5 10 0:12 Nav. camera modif. COTS 0:5 0:5 to 2:0 Laser altimeter modif. COTS 0:05 0:01 Communication TRL 2 1:5 10:0 Total 12:1 kg + 20% Margin w/o unmodif. COTS 13:2 kg Table 3: Mass and power budget for DISCUS. Note that the radar and electric propulsion are not running simultaneously. Power refers to nominal (non-peak) power. The radar power refers to the short stepped- frequency pulse sequences. In our opinion, a good reference candidate for 2021 is ing plane is supposed to avoid shadow phases, which 7 8 the asteroid 65717 (1993 BX3) because of its close would stress the thermal management and decrease approach of 18.4 lunar distances (LD), relative v of the power generation. If the orbit of the spacecraft is around 3:6 km s in January 2021 and diameter be- polar, i.e., perpendicular to the spin of the asteroid, tween 180 and 410 m . As the current discovery rate a single orbit will be sucient to record a backscat- in the recent years has been around 1500 NEAs per tering dataset for a nearly uniformly distributed set year, more candidates could show up in the future. of measurement points enclosing the targeted NEA. The selection of a mission target has to be revisited We assume that a polar or nearly-polar polygonal or- once when the nal launch window has been xed. bit can be achieved with an orbiting direction that is perpendicular to the Solar System's ecliptic plane, since most NEAs are known to have a retrograde spin 4.5 Radar measurement and mapping (La Spina et al., 2004), i.e., the spin orientation is orbit nearly normal to the ecliptic. The orbital movement of the spacecraft will be very slow. That is, the rel- The two CubeSats will approach the targeted aster- ative velocity between the asteroid's surface and the oid to a stable point between 5 km and 10 km alti- spacecraft will be maximally a few meters per second tude and perform their rst radar measurements from and mainly determined by the spin period. Conse- a static position relative to the target. The mov- quently, Doppler e ects can be omitted. The exact duration of the measurements will depend on several https://ssd.jpl.nasa.gov/sbdb.cgi?sstr=65717 parameters such as the asteroids spin period and or- http://www.minorplanetcenter.net/db_search/show_ object?utf8=&object_id=65717 biting distance. For example, a few weeks might be This size range is based on assumed geometric albedo p of a suciently long period. 0.05<p <0.25. There is an observation by Spitzer Space Tele- For positioning, we will apply the data provided by scope which gives, after radiometric analysis using techniques by Trilling et al. (2016), D=91 m and p =0:72 0:33 the radar and laser altimeter, star trackers, and the (http://nearearthobjects.nau.edu/). However, this albedo optical camera. The goal is to obtain an orientation solution is unusually high as largest plausible albedos of NEOs accuracy of at least 3 degrees relative to the target are usually not much larger than 0.5 (Trilling et al., 2010). asteroid's surface, which has been suggested to be suf- Therefore, we consider this result as a lower limit to the size of 65717 (1993 BX ). cient for tomography in (Takala et al., 2017c). The 11 Figure 5: Potential targets in the 2010{2022 time frame. The black outlined box shows the most interesting target range with respect to the diameter and the distance of the closest approach. The gray box shows the asteroid 65717 (1993 B3) which is currently regarded as the most promising candidate with respect to the diameter (180 to 410 m), closest approach distance (18.4 lunar distances) and also v (3:6 km s ). 12 asteroid will be visible in the star trackers providing tions capable to receive in the 31:8 to 32:3 GHz band. an accuracy around 0.01 degrees (Enright, 2010), e.g., With a 3 W transmitter RF-power we expect that a with BST's iADCS100. The distance to the asteroid data rate of about 520 Kibibit=s can be achieved us- can be obtained based on the altimeter data. Using ing the DVB-S2 QPSK 2/3 operations mode. With the light-curve inversion, a preliminary estimate for this transfer rate a dataset of the estimated size can the spin and shape model can be determined dur- be transmitted to the Earth within less than 7 hours. ing the approach with an estimated couple of degrees accuracy for the spin (Durech et al., 2015). As sug- gested by our present analysis, the shape model can 5 Discussion be re ned up to an angular accuracy around one de- In the DISCUS concept, a bistatic (two-spacecraft) gree using the radar at the orbit (Section 3.2). A low-frequency radar is carried by two CubeSats. The further re nement will be attempted via optical im- primary goal is to resolve the interior structure of a ages. 260 to 600 m diameter, i.e., Itokawa-size (Abe et al., In order to ensure that a robust tomographic re- 2006), rubble pile asteroid. Based on this study, the construction can be produced, we use a 1.5{2.0 spa- most promising target candidate is currently 65717 tial oversampling rate with respect to the Nyquist (1993 BX3), a potentially hazardous object, which criterion (NC) which was suggested for a monochro- will make its next close Earth approach in 2021 at matic (travel-time) measurement in (Pursiainen and a v of 3:6 km s with the closest point being at Kaasalainen, 2015). This means recording the radar 18.4 lunar distances from the Earth. More candidates signal for 5000 to 20 000 measurement positions, will probably be detected in the near future, and when extrapolating the measurement positions to the the nal target asteroid can only be selected when close proximity of a targeted 260 to 600 m diame- a launch opportunity has been found. ter NEA. The resulting estimated maximal amount of data recorded by a single satellite will be 250{ Akin to airborne GPR, the radar design is based on 1000 MB, respectively. Two identical CubeSats will the stepped-frequency measurement technique cou- be present in the mission. A perpendicular orbit pled with a linear dipole antenna (Fu et al., 2014). will be used for both satellites. The measurement is We aim to use a low 20{50 MHz signal frequency bistatic, that is, one CubeSat both transmits and re- in order to maximize the signal penetration depth (Francke and Utsi, 2009; Leucci, 2008; Kofman, 2012) ceives the signal and the other one serves as an addi- and, thereby, to obtain a global data coverage for tional receiver. In the measurements, the spacecraft the interior part. The main mission objective will will be placed within a constant angular distance be achieved, if the internal macroporosity structures from each other as seen from the asteroid (Takala can be detected and mapped. We expect that full or et al., 2017a). close to full penetration will be achieved, since the level of the signal attenuation in the porous miner- 4.6 Data link als of planetary subsurfaces in the planned low fre- quency range is known to be low (Kofman, 2012), The essential prerequisite for the downlink is the e.g., 10 to 30 dB km for basalt. Also the outcome ability to transfer the whole scienti c dataset dur- of CONSERT shows that signal penetration through ing the visibility to one ground-station on Earth. an small Solar System body of two times the targeted The size of the dataset is estimated to be about 12 to 13 Gibibit, coming from the radar (500 MiB), the size range can be achieved (Kofman et al., 2015). camera (1025 MiB) and an altimeter (50 MiB). The Moreover, the estimated data transfer capacity (Sec- size of the antenna is assumed to be 20 cm 30 cm, tion 4.6) allows recording the signal for a suciently the available transmit power is limited by the size dense distribution of points within targeted size range of the spacecraft's solar array. Our baseline strat- (Section 4.5). egy is to use the 35 m ESA Deep-Space Ground Sta- Our recent simulation study (Takala et al., 2017a) 13 suggests that the tomographic reconstruction of the ing smaller details. Techniques to increase the band- interior details, such as voids, will be feasible. Based width can, in principle, include adding transmission on the results, we expect that the interior can be channels and modifying the antenna folding, load- reconstructed, if the total noise including both mea- ing and increasing the width (Fu et al., 2014; Pog- surement and modeling errors is at least 5 to 10 dB gio and Mayes, 1971; Balanis, 2012; Alexander et al., below the signal amplitude. This was observed to 2002). Nevertheless, each of these adjustments would be the case in CONSERT (comet 67P/Churyumov- increase the complexity, total weight and power con- Gerasimenko), where the secondary signal peaks were sumption of the system and, therefore, those are not mainly 20 dB below the main peak (Kofman et al., suggested for the initial design. Our initial hard- 2015). Due to the relatively high electric permit- ware design relies on optimizing the weight and size tivity of the asteroid materials we expect a some- of the spacecraft in order to minimize the mission what higher error level for an asteroid. The results of costs. Therefore, the 6U frame is used as the refer- (Takala et al., 2017a), furthermore, suggest that also ence design. Based on this study, the 75 W Busek limited-angle measurements following from a non- BIT-3 thruster will be able to provide the 6U Cube- perpendicular (non-polar) orbit will allow obtaining Sat a v of 3:2 km s which together with an ini- a reconstruction. That is, also NEAs with an excep- tial v obtained from a lunar transfer orbit (Landis tional (non-retrograde) spin (La Spina et al., 2004) et al., 2014) will result in a total v of approximately are potential targets for tomographic reconstruction. 4 km s . This would be enough to reach, e.g., the as- In the numerical simulations, we have assumed a teroid 65717 (1993 BX3) in 2021. Depending on the 25 dB km signal attenuation level. In practice, the eccentricity of the transfer orbit the targeted 75 W scattering caused by the macroporosity can lead to cannot be achieved for the entire transfer. Yet, we ex- some additional attenuation. pect that including the initial v, the required power The targeted 2 MHz total bandwidth for the for the thruster will have enough margins to compen- stepped-frequency measurements would give the res- sate these phases. In the current design the total v olution of at least 20 to 40 m inside the target as- is limited mainly by the power provided by the so- teroid (Daniels, 2005), if its relative permittivity is lar cells, as they degrade over time and produce less 3 to 12 (e.g. solid and pulverized rocks and silica). power at sun distances beyond 1 AU. The number This would mean the distinguishability of the details of reachable targets would increase if solar cells with around the size 1=20{1=10 with respect to the as- 120 W are used, as the potential of the propulsion teroid diameter providing an unforeseen reconstruc- system could then be used to the full extend. With tion accuracy for the deep interior structure. Con- currently available solar panels this would the mass sequently, the current design should be sucient for by 0:8 kg. the main goal of detecting major variances in the vol- A larger 12U size can also be considered, even if umetric relative electric permittivity structure. For all components would t in the 6U bus and other au- comparison, the initial maximal resolution estimates thors have described the envelope as feasible (Frick for CONSERT and Deep Interior were 16 and 20 m, et al., 2014; Staehle et al., 2013). Namely, the mar- respectively (Kofman et al., 2007; Asphaug et al., gins are small and could be eaten up by any inser- 2003). In CONSERT, the carrier frequency and tion in the radar, communication system or thermal bandwidth of the signal were 90 and 8 MHz, respec- management. Nevertheless, a 12U bus would lead tively. to a higher total weight and, consequently, require a We consider the 2 MHz bandwidth advantageous much larger orbit control system. As the power would from both computational and measurement view- not necessarily scale with size, a completely di erent points, as we currently can invert a full set of data for propulsion system would be needed. One possibility the targeted asteroid size (Takala et al., 2017a) and to overcome the possible issues would be a piggy-bag sucient measurement accuracy seems achievable. A ride to an asteroid. This would reduce the need of wider bandwidth would, however, allow distinguish- v and communication capability and might lead to 14 an extra 3/2 U of scienti c payload, for instance, a to reconstruct the interior of a comet and also in map- more advanced camera system. ping of the recently discovered Moon caves. Several The proposed self-rolled half-wavelength dipole an- details will need to be re ned in the nal design stage. tenna design is bene cial with respect to the mass and volume. We emphasize that the present an- tenna speci cations have been designed for Itokawa- size bodies for which a well-penetrating signal fre- 6 Conclusion quency is necessary. Nevertheless, the radar and the spacecraft can also operate with higher frequencies DISCUS - The Deep Interior Scanning CubeSat mis- and shorter antenna lengths, which might be prefer- sion is designed with the aim to characterize the in- able for smaller bodies or shallower subsurface struc- terior structure of rubble pile asteroids. Computed tures. In any case COTS solutions at slightly higher radar tomography and a bistatic con guration of two weight are available. identical CubeSats will be used to reconstruct inter- We consider that the goal to t the radar in a 1U nal electric permittivity heterogeneities. While deep- (1000 cm ) space is feasible based on RST's earlier space missions for CubeSat sized spacecraft are still experience of the 900 cm HOPE radar design (RST a challenge from an engineering point of view, we Radar Systemtechnik AG, 2001). The next devel- found that our radar is compatible with the limita- opment stage aim to build a 20 MHz radar demon- tions of a CubeSat mission to a near-Earth asteroid. strator to validate the proposed instrumentation in The baseline design of the DISCUS spacecraft pro- an on-ground test. RST GmbH has previous experi- vides a v of 3:2 km s . A launch to a lunar trans- ence of such an experiment from ESA's project PIRA fer orbit provides an additional v so that asteroids (Braun et al., 1997) in which a stepped-frequency with a v of approximately 4 km s are reachable. radar equipped with a sub 10 MHz orbiter antenna The most prominent target candidate among the cur- system was successfully tested under motion. An on- rently known NEAs is 65 717 (1993 BX3). going work is to transfer the existing radar design to a below 100 MHz frequency band. The main scienti c result that could be accom- Compared to more frequently used microwave plished with DISCUS mission is mapping the distri- CubeSat radars, for example, Raincube's 4U and bution of porosity within a rubble pile asteroid. This 10 W Ka-band radar (Peral et al., 2017), the planned would be on the one hand of great value to formation low-frequency band allows a more simple design. At- models of the Solar System, but also of great impor- taining a considerably better synchronization than tance to de ection scenarios for potentially hazardous what is natively provided by the CSAC, e.g., due to asteroids imposing an impact threat to earth. Addi- a wider bandwidth, might necessitate extra arrange- tionally, crucial knowledge of the internal permittiv- ments. For example, in the recent microwave-range ity structures and, thereby, also the mineral content interferometry measurements of the TandDEM-X of the asteroids would be obtained. mission, obtaining a picosecond-level synchroniza- tion required relative phase referencing via six ded- icated synchronization horn antennas in each space- craft (Krieger et al., 2007). Acknowledgments Future work will include construction of a demon- strator prototype to validate the radar. Numerical simulations concerning the stepped-frequency mea- MT and SP were supported by the Academy of Fin- surement technique, shape modeling, and also orbit land Key Project 305055 and AoF Centre of Excel- calculations around the asteroid will be performed. lence in Inverse Problems. We will also examine the capability of the planned PB, JD and EV were supported by the Max Planck instrumentation, e.g., regarding yby measurements, Institute for Solar System research. 15 References Cube Mission: Determining Lunar Water Dynam- ics with a First Generation Deep Space CubeSat. Abe, S., Mukai, T., Hirata, N., Barnouin-Jha, Science 330, 463{468. O. S., Cheng, A. F., Demura, H., Gaskell, R. W., Daniels, D. J., 2005. Ground penetrating radar. Wi- Hashimoto, T., Hiraoka, K., Honda, T., Kubota, ley Online Library. T., Matsuoka, M., Mizuno, T., Nakamura, R., Scheeres, D. J., Yoshikawa, M., 2006. Mass and Deller, J. F., 2017. Hyper-Velocity Impacts on Rub- Local Topography Measurements of Itokawa by ble Pile Asteroids. Ph.D. thesis, School of Physical Hayabusa. Science 312 (5778), 1344{1347. Sciences University of Kent Canterbury U.K. and Alexander, M., Salter, M., Loader, B., Knight, D., International Max Planck Research School for So- lar System Science G ottingen Germany. 2002. Broadband calculable dipole reference anten- nas. IEEE Transactions on electromagnetic com- DeNatale, J., Borwick, R., Tsai, C., Stupar, P., patibility 44 (1), 45{58. Lin, Y., Newgard, R., Berquist, R., Zhu, M., Asmar, S., Matousek, S., 2014. Mars Cube one 2008. Compact, low-power chip-scale atomic clock. (MarCO): The rst planetary CubeSat mission. In: In: Position, Location and Navigation Symposium, Proceedings of the Mars CubeSat/NanoSat Work- 2008 IEEE/ION. IEEE, pp. 67{70. shop, Pasadena, California, November. Vol. 20. Durech, J., Carry, B., Delbo, M., Kaasalainen, M., p. 21. Viikinkoski, M., 2015. Asteroid models from mul- Asphaug, E., Belton, M., Cangahuala, A., Keith, L., tiple data sources. Asteroids IV, 183. Klaasen, K., McFadden, L., Neumann, G., Ostro, Eisenburger, D., Lentz, H., Jenett, M., 2008. S., Reinert, R., Safaeinili, A., et al., 2003. Explor- Helicopter-borne GPR systems: A way from ice ing asteroid interiors: The Deep Interior mission thickness measurements to geological applications. concept. In: Lunar and Planetary Science Confer- In: Ultra-Wideband, 2008. ICUWB 2008. IEEE In- ence. Vol. 34. ternational Conference on. Vol. 3. IEEE, pp. 161{ Balanis, C., 2012. Antenna Theory: Analysis and De- 165. sign. Wiley. Enright, J., 2010. Moon-tracking modes for star Barnouin-Jha, O. S., Cheng, A. F., Mukai, T., Abe, trackers. Journal of guidance, control, and dynam- S., Hirata, N., Nakamura, R., Gaskell, R. W., ics 33 (1), 171{185. Saito, J., Clark, B. E., 2008. Small-scale topog- Erst, S. J., 1984. Receiving systems design. Artech raphy of 25143 Itokawa from the Hayabusa laser House on Demand, Dedham, Massachusetts. altimeter. Icarus 198 (1), 108 { 124. Farinella, P., Paolicchi, P., Zappal a, V., Dec. 1982. Braun, H., Lentz, H., Woode, A., 1997. A Planetary The asteroids as outcomes of catastrophic colli- Into-the-Ground Radar and Altimeter (PIRA). In: sions. Icarus 52, 409{433. Applied Geologic Remote Sensing - International Conference. Vol. 1. pp. I{498. Folta, D. C., Bosanac, N., Cox, A., Howell, K. C., 2016. The Lunar IceCube mission design: con- Carry, B., 2012. Density of asteroids. Planetary and struction of feasible transfer trajectories with a Space Science 73 (1), 98{118. constrained departure. NASA Technical Reports Clark, P., Malphrus, B., Brown, K., Reuter, D., Mac- Server (AAS 16-285). Dowall, R., Folta, D., Mandell, A., Hurford, T., URL https://ntrs.nasa.gov/archive/nasa/ Brambora, C., Patel, D., et al., 2016. Lunar Ice casi.ntrs.nasa.gov/20170001470.pdf 16 Francke, J., Utsi, V., 2009. Advances in long-range Jol, H. M., 2008. Ground Penetrating Radar Theory GPR systems and their applications to mineral ex- and Applications. Elsevier Science. ploration, geotechnical and static correction prob- Kofman, W., 2012. Radar techniques to study sub- lems. First Break 27 (7). surfaces and interiors of the solar system objects. Frick, A., Castillo-Rogez, J., Johnson, L., Dervan, In: Microwave Radar and Wireless Communica- J., 2014. NEA Scout: A CubeSat Architecture for tions (MIKON), 2012 19th International Confer- Near Earth Easteroid (NEA) Exploration. In: Pro- ence on. Vol. 2. IEEE, pp. 409{412. ceedings of Interplanetary Small Satellite Confer- Kofman, W., Herique, A., Barbin, Y., Barriot, J.-P., ence (ISSC). Ciarletti, V., Cli ord, S., Edenhofer, P., Elachi, C., Fu, L., Liu, S., Liu, L., Lei, L., 2014. Development Eyraud, C., Goutail, J.-P., Heggy, E., Jorda, L., of an airborne ground penetrating radar system: Lasue, J., Levasseur-Regourd, A.-C., Nielsen, E., antenna design, laboratory experiment, and nu- Pasquero, P., Preusker, F., Puget, P., Plettemeier, merical simulation. IEEE Journal of selected topics D., Rogez, Y., Sierks, H., Statz, C., Svedhem, H., in applied Earth observations and remote sensing Williams, I., Zine, S., Van Zyl, J., 2015. Proper- 7 (3), 761{766. ties of the 67P/Churyumov-Gerasimenko interior revealed by CONSERT radar. Science 349 (6247). Gundelach, V., Blindow, N., Buschmann, U., Salat, C., Krellmann, Y., 2010. Exploration of geologi- Kofman, W., Herique, A., Goutail, J.-P., Hagfors, T., cal structures with GPR from helicopter and on Williams, I. P., Nielsen, E., Barriot, J.-P., Barbin, the ground in the Letzlinger Heide (Germany). In: Y., Elachi, C., Edenhofer, P., Levasseur-Regourd, Ground Penetrating Radar (GPR), 2010 13th In- A.-C., Plettemeier, D., Picardi, G., Seu, R., Sved- ternational Conference on. IEEE, pp. 1{6. hem, V., 2007. The comet nucleus sounding exper- iment by radiowave transmission (CONSERT): A Herique, A., Ciarletti, V., 2016. A Direct Observation short description of the instrument and of the com- of the Asteroids Structure from Deep Interior to missioning stages. Space Science Reviews 128 (1{ Regolith: Two Radars on the AIM Mission. In: 4), 413{432. 47th Lunar and Planetary Science Conference. p. Kraus, J., 1966. Radio astronomy. McGraw-Hill. Hodges, R. E., Chahat, N. E., Hoppe, D. J., Vac- Krieger, G., Moreira, A., Fiedler, H., Hajnsek, I., chione, J. D., 2016. The Mars Cube One deployable Werner, M., Younis, M., Zink, M., 2007. TanDEM- high gain antenna. In: Antennas and Propagation X: A satellite formation for high-resolution SAR (APSURSI), 2016 IEEE International Symposium interferometry. IEEE Transactions on Geoscience on. IEEE, pp. 1533{1534. and Remote Sensing 45 (11), 3317{3341. Imken, T., Castillo-Rogez, J., He, Y., Baker, J., Mari- nan, A., 2017. Cubesat ight system development Kwiatkowski, T., Jan. 2010. Photometric survey of for enabling deep space science. In: Aerospace Con- the very small near-Earth asteroids with the SALT ference, 2017 IEEE. IEEE, pp. 1{14. telescope. Astronomy and Astrophysics 509, A95. Ishiguro, M., Nakamura, R., Tholen, D. J., Hirata, La Spina, A., Paolicchi, P., Kryszczynska, A., Pravec, N., Demura, H., Nemoto, E., Nakamura, A. M., P., 2004. Retrograde spins of near-Earth asteroids Higuchi, Y., Sogame, A., Yamamoto, A., Kitazato, from the Yarkovsky e ect. Nature 428 (6981), 400. K., Yokota, Y., Kubota, T., Hashimoto, T., Saito, J., Jun. 2010. The Hayabusa Spacecraft Asteroid Lacomme, P., 2001. Air and Spaceborne Radar Sys- Multi-band Imaging Camera (AMICA). Icarus 207, tems: An Introduction. Electronics & Electrical. 714{731. William Andrew. 17 Landis, G. A., Oleson, S. R., McGuire, M. L., Burke, of boulders on global surface of asteroid 25143 L. M., Martini, M. C., Fittje, J. E., Packard, T. W., Itokawa. In: Earth Planets and Space. Fukushima 2014. A cubesat asteroid mission: Propulsion Natl Coll Technol, Iwaki, Fukushima 9708034, trade-o s. In: 50th AIAA/ASME/SAE/ASEE Japan, pp. 13{20. Joint Propulsion Conference. AIAA Propulsion Nakamura, T., Nakamura, A. M., Saito, J., Sasaki, S., and Energy Forum. American Institute of Aero- Nakamura, R., Demura, H., Akiyama, H., Tholen, nautics and Astronautics. D., et al., 2001. Multi-band imaging camera and its Lentz, H., Braun, H., 2000. GINGER II An up- sciences for the Japanese Near-Earth asteroid mis- grade of the technology demonstrator of the guid- sion MUSES-C. Earth, Planets and Space 53 (11), ance and into the ground exploration radar Gin- 1047{1063. ger. Advanced Space Technologies for Robotics and NASA Mission Design Division, 2015. Small Automation (ASTRA 2000), ESTEC, Noordwijk, spacecraft technology state of the art. Holland, paper (3.5). https://www.nasa.gov/sites/default/files/ atoms/files/small_spacecraft_technology_ Leucci, G., 2008. Ground Penetrating Radar: The state_of_the_art_2015_tagged.pdf, accessed: Electromagnetic Signal Attenuation and Maximum 2017-10-05. Penetration Depth. Scholarly Research Exchange Peral, E., Statham, S., Taneli, S., Imken, T., Williams, A., Price, D., Sauder, J., Chahat, N., Manasse, R., 1960. Summary of maximum theoret- 2017. RainCube, a Ka-band Precipitation Radar in ical accuracy of radar measurements. Tech. rep., a 6U CubeSat. In: 31st Annual AIAA/USU Con- MITRE CORP BEDFORD MA. ference on Small Satellites. Michel, P., DeMeo, F., Bottke, W., 2015. Asteroids Poggio, A., Mayes, P., 1971. Bandwidth extension IV. Space Science Series. University of Arizona for dipole antennas by conjugate reactance loading. Press. IEEE Transactions on Antennas and Propagation Michel, P., Kueppers, M., Carnelli, I., Galvez, A., 19 (4), 544{547. Mellab, K., Cheng, A., 2016. Asteroid Impact Mis- Pravec, P., Harris, A. W., Nov. 2000. Fast and Slow sion (AIM): The European component of the AIDA Rotation of Asteroids. Icarus 148, 12{20. space project. In: Lunar and Planetary Science Conference. Vol. 47. p. 1204. Pravec, P., Harris, A. W., Michalowski, T., 2002. As- teroid rotations. Asteroids III 113. Michel, P. P., Benz, W. W., Tanga, P., Richard- son, D. C., Nov. 2001. Collisions and Gravitational Pursiainen, S., Kaasalainen, M., 2015. Electromag- Reaccumulation: Forming Asteroid Families and netic 3D subsurface imaging with source sparsity Satellites. Science 294 (5547), 1696{1700. for a synthetic object. Inverse Problems 31 (12), Michel, P. P., Benz, W. W., Tanga, P., Richardson, D. C., Nov. 2002. Formation of Asteroid Fami- Putz, P., Woode, A., Braun, H.-M., 1997. Ginger - a lies by Catastrophic Disruption: Simulations with radar instrument for planetary rover applications. Fragmentation and Gravitational Reaccumulation. Preparing for the Future 7 (2), 14. Icarus 160 (1), 10{23. Rahmat-Samii, Y., Manohar, V., Kovitz, J. M., 2017. Michikami, T., Nakamura, A. M., Hirata, N., Gaskell, For satellites, think small, dream big: A review of R. W., Nakamura, R., Honda, T., Honda, C., recent antenna developments for cubesats. IEEE Hiraoka, K., Saito, J., Demura, H., Ishiguro, Antennas and Propagation Magazine 59 (2), 22{ M., Miyamoto, H., 2008. Size-frequency statistics 30. 18 Reveles, J., Lawton, M., Fraux, V., Gurusamy, V., Snodgrass, C., Jones, G., Boehnhardt, H., Gibbings, Parry, V., 2017. In-orbit performance of astrotube: A., Homeister, M., Andre, N., Beck, P., Bentley, Alsat nano's low mass deployable composite boom M., Bertini, I., Bowles, N., Capria, M., Carr, C., payload. Ceriotti, M., Coates, A., Corte, V. D., Hanna, K. D., Fitzsimmons, A., Gutirrez, P., Hainaut, O., Richardson, D. C., Leinhardt, Z. M., Melosh, H. J., Herique, A., Hilchenbach, M., Hsieh, H., Jehin, Bottke, W. F., Asphaug, E., Oct. 2002. Gravi- E., Karatekin, O., Kofman, W., Lara, L., Laudan, tational Aggregates: Evidence and Evolution. In: K., Licandro, J., Lowry, S., Marzari, F., Masters, Bottke, W. F., Cellino, A., Paolicchi, P., Binzel, A., Meech, K., Moreno, F., Morse, A., Orosei, R., R. P. (Eds.), Asteroids III. University of Arizona Pack, A., Plettemeier, D., Prialnik, D., Rotundi, Press, pp. 501{515. A., Rubin, M., Snchez, J., Sheridan, S., Trielo , M., Winterboer, A., 2017. The Castalia mission to Rozitis, B., MacLennan, E., Emery, J. P., Aug. 2014. Main Belt Comet 133P/Elst-Pizarro. Advances in Cohesive forces prevent the rotational breakup Space Research. of rubble-pile asteroid (29075) 1950 DA. Nature 512 (7513), 174{176. Staehle, R., Blaney, D., Hemmati, H., Lo, M., RST Radar Systemtechnik AG, 2001. HOPE (Hand- Mouroulis, P., Pingree, P., Wilson, T., Puig-Suari, held Operated Demining System). http://www. J., Williams, A., Betts, B., et al., 2013. Interplane- rst-group.biz/index.php?id=113&no_cache=1, tary CubeSats: opening the solar system to a broad accessed: 2017-10-05. community at lower cost. Journal of small satellites 2 (1), 161{186. Rutishauser, A., Maurer, H., Bauder, A., 2016. Helicopter-borne ground-penetrating radar investi- Takala, M., Bambach, P., Deller, J., Vilenus, E., Wit- gations on temperate alpine glaciers: A comparison tig, M., Lentz, H., Braun, H. M., Kaasalainen, of di erent systems and their abilities for bedrock M., Pursiainen, S., 2017a. A far- eld inversion mapping. Geophysics 81 (1), WA119{WA129. approach for the deep interior scanning cubesat. arXiv preprint arXiv:1709.04309. Safaeinili, A., Gulkis, S., Hofstadter, M. D., Jordan, R. L., Dec. 2002. Probing the interior of asteroids Takala, M., H am al ainen, T. D., Pursiainen, S., and comets using radio re ection tomography. Me- 2017b. The E ect of Hardware-Computed Travel teoritics and Planetary Science 37, 1953{1963. Time on Localization Accuracy in the Inversion Saito, J., Miyamoto, H., Nakamura, R., Ishiguro, of Experimental (Acoustic) Waveform Data. IEEE M., Michikami, T., Nakamura, A. M., Demura, Transactions on Computational Imaging 3 (2), H., Sasaki, S., Hirata, N., Honda, C., Yamamoto, 344{354. A., Yokota, Y., Fuse, T., Yoshida, F., Tholen, Takala, M., Us, D., Pursiainen, S., 2017c. Multigrid- D. J., Gaskell, R. W., Hashimoto, T., Kubota, T., based inversion for volumetric radar imaging with Higuchi, Y., Nakamura, T., Smith, P., Hiraoka, K., asteroid interior reconstruction as a potential ap- Honda, T., Kobayashi, S., Furuya, M., Matsumoto, plication. arXiv preprint arXiv:1707.04605. N., Nemoto, E., Yukishita, A., Kitazato, K., Der- mawan, B., Sogame, A., Terazono, J., Shinohara, Thirouin, A., Moskovitz, N., Binzel, R. P., Chris- C., Akiyama, H., 2006. Detailed Images of Asteroid tensen, E., DeMeo, F. E., Person, M. J., Polishook, 25143 Itokawa from Hayabusa. Science 312 (5778), D., Thomas, C. A., Trilling, D., Willman, M., Hin- 1341{1344. kle, M., Burt, B., Avner, D., Aceituno, F. J., Nov. S anchez, P., Scheeres, D. J., Apr. 2012. DEM simula- 2016. The mission accessible Near-Earth Objects tion of rotation-induced reshaping and disruption survey (MANOS): First photometric results. The of rubble-pile asteroids. Icarus 218 (2), 876{894. Astronomical Journal 152 (6), 163. 19 Tricarico, P., Mar. 2017. The near-Earth aster- oid population from two decades of observations. Icarus 284, 416{423. Trilling, D. E., Mommert, M., Hora, J., Chesley, S., Emery, J., Fazio, G., Harris, A., Mueller, M., Smith, H., Dec. 2016. NEOSurvey 1: Initial Re- sults from the Warm Spitzer Exploration Science Survey of Near-Earth Object Properties. Astro- nomical Journal 152, 172. Trilling, D. E., Mueller, M., Hora, J. L., Harris, A. W., Bhattacharya, B., Bottke, W. F., Chesley, S., Delbo, M., Emery, J. P., Fazio, G., Mainzer, A., Penprase, B., Smith, H. A., Spahr, T. B., Stans- berry, J. A., Thomas, C. A., Sep. 2010. Explore- NEOs. I. Description and First Results from the Warm Spitzer Near-Earth Object Survey. Astro- nomical Journal 140, 770{784. Walker, R., Koschny, D., Bramati, C., Carnelli, I., ESA, C. S. T., 2017. Miniaturised Asteroid Re- mote Geophysical Observer (M-ARGO): a stand- alone deep space CubeSat system for low-cost sci- ence and exploration missions. In: iCubeSat 7 ab- stracts. Walsh, K. J., Richardson, D. C., Michel, P. P., Jul. 2008. Rotational breakup as the origin of small bi- nary asteroids. Nature 454 (7201), 188{191.
Journal
Astrophysics – arXiv (Cornell University)
Published: May 4, 2018